I’m having a little trouble understanding how the DCC auto reverse works after the loco re-enters the main. I bought two PSX-ARs for two loops, but what about the main. Don’t I have to reverse the main as well? I had to reverse the main on my DC layout as the loco came out of the loop and I don't see any mention of it in DCC. How does that work?
The difference is that on DCC, the direction the engine runs does not depend on track polarity. As soon as the first metal wheelset hits the gap between the loop and main (at either end of the loop), the PSX-AR simply matches the loop polarity to the main polarity. Once the engine is back on the main, the decoder doesn't care what the track polarity is. It will continue to go in the same direction. In theory, you could hook the PSX-AR to either the loop or the main, but there's no need to do both.
That’s interesting, so why do I need it if the AR reverses the polarity of the loop when entering, but the loco can exit to the main when the main is now the opposite polarity without a problem? It seems like something more has to happen.
I was confused also by how the reversing module works also. I put in a Digitrax AR1 on a Wye on my clubs layout. Apparently the reversing is done in the decoder as you can have a train on the main as well as one in the reversing section and the one on the main does not change direction.
Don't forget, the AR always reverses the polarity of the loop when the engine is leaving the loop, also as required when it enters the loop. The important thing is that the polarity must match while the engine is passing over the gap. If all your trains go around the loop clockwise, for example, the AR will trigger twice; once when the train enters the loop, and once when it leaves.
I guess where I’m confused is that it IS the same scenario whether the loco is entering or exiting. In both scenarios its “loco enters track of opposite polarity”. Now maybe I don’t fully understand, but I thought the auto reverse simply reverses track polarity. I don’t understand why the decoder can’t handle the polarity in one situation and not the other. In the 1st scenario the polarity of the track being entered is reversed to match the track which the loco is coming from. In the 2nd scenario the track being entered is not reversing polarity to match the track which the loco is coming from. I don’t know what I’m missing and I do understand that it works although I have not witnessed it yet. So what is the difference, what am I missing here?
I guess I was posting at the same time. So I can probably understand it if the the loop track reverses back before the loco exists.
Ok now that I think about it, that makes more sense. I did not know that it worked like that. I only knew how the Circuitron AR-1 auto reversing unit worked for DC operation. Now I understand why the Digitrax AR1 doesn’t work on DC. Thanks for Clearing that up. :thumbs_up:
One other question about reversing block length I choose the PSX-AR because it has a Block Occupied output that I plan on using to switch a turnout through my DS64. I’m trying to figure out where to block off my reversing loop so that it the turnout will not switch too soon. From everything I’ve read, the block needs to be as long as the longest train. I’m now wondering what is going to happen if I narrow that up to keep the turnout open to outside traffic until the train is actually close enough to the turnout. How short can I make this reversing block without other problems and how does it know how long the train is? There are only two wires connected to the track and I would think that as each Car wheel crosses the gaps there would be a small short. Is that going to cause the AR to constantly flip back and forth until the last car crosses over.
The reverser module only triggers once in each reverse loop usage cycle, not twice. If it triggered twice, it would create its own short on either the exit or the entrance. Instead, the reverser senses each gap crossing and looks only for a mismatch. If there is no mismatch, it does nothing. That may be on either end of the loop. For example, the reverser detects no polarity disparity when the train enters the loop. It does nothing. On exit, logically, there must be a disparity, and it effects the necessary change in polarity. Or, it senses a mismatch when the engine crosses the first gap, and effects the change. If it had to change initially, it won't at the far end, and nothing will happen there since on that end the polarity matches.
Ok, so if it only switches once and it doesn’t matter if the loco is entering or exiting, then I’m assuming the only reason for the switch in polarity is to prevent a short and nothing more. A DC Auto reverser has two functions, one function is to prevent a short and the other is to keep the loco moving in one direction. I think I got it!
I think I wasted my money by buying two PSX-ARs because I already have a PM42 which will handle both loops. The only thing I needed was a BD4 for block detection. Maybe I can return the PSX-AR.
It doesn't really matter but that isn't correct. The nature of the reverse loop is such that whichever way the AR switches one end will have reversed polarity to the adjacent 'main'. If on entering the loop a reversal is needed to correct the short, this happens. But this means that the far end will now definitely be the wrong polarity and the AR unit will change it again when the engine reaches that point - which automatically means the entry point is now reversed polarity. So for a loop that is always traversed in the same direction (eg. using a spring switch) the reverser will always operate twice - on entry and on exit. Yes. And putting these two points together is the reason the loop should be longer than any train, otherwise you could have locos and metal wheeled cars bridging both gaps at the same time, which means the AR cannot adopt a correct polarity. (In fact you will also have a short of the main controller too, because the two ends of the main are effectively connected together.)
An auto reverser senses a short circuit the same way as a circuit breaker. What it does when the short is sensed is the difference between it and a circuit breaker. A circuit breaker will kill power when it senses a short and restores it when the short is no longer present. An auto reverser simply swaps the output polarity (phase) when it senses a short circuit. If the short continues it will keep swapping the outputs. An AR1 actually will work with DC but it's not much use with DC (I tried it) because it flips polarity under the train which will reeverse the train and send it backwards. I'm pretty sure that a PSX will not work at all on DC but never tried. An AR1 or PM42 auto-reverser might let a DC engine enter a reverse loop but it would never get back out. Unlike DCC, a DC reverse loop requires the next block to match polarity before the train can continue. Martin Myers
Martin is right. In fact, The way the PM42 is configured takes advantage of this. For circuit breaker use, the normally closed set of relay contacts is used, which open up when a short is detected. For autorevese use, the normally open relay contacts (which are brought out onto the edge connector and are wired to switch polarity) are also used (There are slightly different sensing parameters for the two modes as set by the OpSws as well).
After reviewing all this great info, I want to try just using the PM42 without the PSX-AR to see how it will work. Unfortunately I’m still waiting for a replacement power supply for my DCC system. This brings up another question. All the examples show a booster connected to the PM42 as well as the PSX-AR. Am I going to be able to just use my NCE power supply to operate this?
Johnny, Give this a read. It's the simplest thing I've found yet. It also agrees with the information and answers given here. Test Results - MRC Auto-Reverser - December 1998 update | Tech News at Tony's Train Exchange.
Thanks Rick. I think I got it figured out. I think what I was missing was that the DCC loco did not care what the polarity was and could handle the change in polarity in the track it was on as long as it wasn’t a short where as the straight DC loco needed the correct polarity to continue in the correct direction. I could not understand why you had to change polarity of the loop thru the 2PDT relay on the circuit board to enter but not need to change the polarity of track it was exiting to as in DC. Now I know that once the train is in the loop, the polarity can change without any effect to the operation of the loco, so the polarity changes on the loop to match the track that the loco is entering and there isn’t a short. I was concerned about what was going to happen to the Decoder exiting the loop and what was going to happen to other locos already running on the main if I changed polarity. Now I know that it doesn’t mater as long as it isn’t a short. I had a similar debate on another forum about the AR1 because I knew about the 2PDT relay on board that was no different than a 2PDT toggle switch and I theorized that as long something changed the polarity of the track ahead, whether it was a Toggle switch or another AR1, I figured it would work. I never had two AR1s to confirm my theory, but I read in another thread about a guy who had used an AR1 for DC but had to use a toggle Switch for changing the polarity of the track ahead. Two Circuitron AR-1s does that in DC.
